From the Labs: Information Technology

Building a Nano Radio
A radio receiver made from a carbon nanotube could be used to wirelessly transmit data from ultrasmall sensors

Source: “Nanotube Radio”
Alex Zettl et al.
Nano Letters 7: 3508-3511

Results: Researchers at Lawrence Berkeley National Laboratory in California have modified a carbon nanotube so that it performs the functions of a radio, even tuning in the entire FM radio band.

Why it matters: Radios are used in everything from cell phones to nodes in sensor networks, and like other electronics, they are shrinking in size. A nanoscale radio could someday find its way into portable electronics such as cell phones. The researchers also suspect that with its small size, the radio could be inserted into a biological cell to transmit information collected by tiny sensors that detect molecular processes.

Methods: The researchers grew the carbon nanotube on a tungsten surface that acts as a negative electrode; a positive copper electrode is separated from the nanotube by a vacuum gap. A voltage applied to the electrodes causes a current to flow through the nanotube, turning the radio on. Changing the voltage also changes the vibrational rate of the nanotube, tuning it to a different frequency.

Next steps: The researchers are looking to integrate the radio into biological systems.

Fixing Bugs in Hardware
Software diagnoses problems in chip prototypes and offers fast, cheap solutions

Source: “Automatic Post-Silicon Debugging and Repair”
Valeria Bertacco et al.
International Conference on Computer-Aided Design, November 6, 2007, San Jose, CA

Results: Researchers at the University of Michigan have developed software that finds flaws in computer chips and proposes economical fixes. The software is able to repair about 70 percent of bugs.

Why it matters: Before a chip is mass-produced, a proto­type is shipped from the fabrication facility to the chip designers for testing. Currently, engineers can spend up to a year manually inspecting a prototype for mistakes, such as design errors, misplaced transistors, or wires that are too close together. Each time flaws are identified and corrected, a new prototype has to be made and tested. Each iteration can cost millions of dollars, and repeated prototyping delays commercialization. Manual bug-hunting is also prone to error and may result in products with faults that can be exploited by computer viruses.

Methods: Engineers test chip prototypes by hooking them up to probes that send electrical stimuli through them and record the output. The Michigan researchers wrote software that quickly runs through thousands of input signals and analyzes the output, zeroing in on problem areas. Likewise, it identifies ways to fix bugs by running through a series of simulations to find a design variation that offers the fastest and most cost-effective solution, one that may not be obvious to an engineer looking at a wiring diagram.

Next steps: The researchers are looking into some of the debugging challenges specific to multicore processors–chips with more than one processing center.